Below: the slide order with what to put on each, how to show the paper’s results and its limitation honestly, how the ending differs for B.Tech and M.Tech, the questions reviewers ask, and a free PowerPoint template at the top of this page.
- Explain the paper, then your plan. Reviewers check that you understand the paper before they approve what you will add.
- Use the paper’s own numbers, with the dataset and the metric named, and never claim results you have not run.
- One limitation, clearly shown, is worth more than a list of vague weaknesses.
- End with something measurable: what you will change and how you will compare it with the paper.
- Prepare the paper’s basics: authors, journal or conference, year and dataset, because the first questions are usually about them.
What is a base paper presentation?
It is the review, often the first or “zeroth” review of a final year or M.Tech project, where you present the base paper your project will reproduce and improve. The panel wants to see three things: that the paper is real and suitable, that you understand how it works, and that your planned change is small enough to finish and clear enough to measure. It comes before the synopsis or alongside it, and the slides you make now are reused in every later review.
What slides should a base paper presentation have?
| Slide | What to show | What the panel checks |
|---|---|---|
| 1. Title | Your project title, team, guide, and the base paper’s title, authors, venue and year | That the paper is a real, published one |
| 2. Problem | The problem in plain words, and who has it | That you can explain why it matters |
| 3. Base paper at a glance | Venue, year, DOI, the paper’s aim in one sentence | A recent, peer-reviewed source |
| 4. Related work | Three or four earlier approaches and where the base paper fits | That you read beyond one paper |
| 5. Method | The paper’s block diagram or pipeline, redrawn by you | Understanding, not a pasted figure |
| 6. Dataset and setup | The dataset, its size and source, the split, the metric | That you can actually get the data |
| 7. Results | The paper’s results table, with your reproduced number if you have one | Honest reporting of what was measured |
| 8. Limitations | One limitation you will address, with evidence from the paper | A real gap, not a guess |
| 9. Proposed work | Your change, and the comparison you will run | Small enough to finish, clear enough to measure |
| 10. Plan | Tools, hardware if any, and phases matched to your review dates | A timeline that leaves room for testing |
| 11. References | The base paper first, then the related work, in IEEE style | Complete, correctly formatted citations |
| 12. Thank you | Your names and a line inviting questions | A clean finish |
Follow your department’s template where it sets one. The free PowerPoint template at the top of this page has these twelve slides with guidance on each.
How do you present the base paper’s results and limitations?
Put the paper’s main results table on one slide and circle the number your project cares about. Say which dataset and which metric it is, in one sentence: for example, “accuracy of 94% on the public test set of this dataset”. If you have reproduced the result, show your number next to the paper’s; a small gap is normal, and explaining it (a different split, fewer training epochs, a smaller GPU) shows the panel you ran it yourself. If you have not run it yet, say so and say when you will.
For the limitation, choose one you can test. Good ones come from the paper itself: a future-work line, a single dataset, results only on clean data, a slow model or no explanation of the output. Weak ones are opinions such as “the accuracy can be improved”. Each limitation should point to the change on the next slide.
How do you present your proposed work?
- M.Tech: one measurable change to the paper’s method, such as a new dataset, a different backbone, a robustness test or a lighter model, and a comparison table you will fill in: the paper’s number, your reproduced number and your improved number, on the same metric. That table becomes the core of your M.Tech thesis.
- B.Tech: the system you will build from the paper: what the user sees, the modules, and how you will show it working at the final review. A reproduced result and a working interface are usually enough.
Need a paper to start from? Our list of IEEE base papers has 28 recent open-access papers with what each suits, and where to find a base paper shows how to search for your own topic.
What questions do reviewers ask about a base paper?
- Who wrote the paper, where was it published and in which year?
- Why did you choose this paper over others in the same area?
- Explain the method on the board, step by step.
- Which dataset did the authors use, and can you get it?
- What does this metric measure, and why did the authors choose it?
- Have you run their code? What did you get?
- What is the main limitation, and how will your change address it?
- How will you know your version is better?
- What will you show working at the final review?
For the full list of viva questions across all reviews, see our project viva questions.
What mistakes weaken a base paper presentation?
- Pasting the paper’s figures and text. Redraw the diagram and explain it in your own words.
- Claiming the paper’s results as your own. Label every number with its source.
- A paper you cannot reproduce. Check the dataset and code are available before the review.
- An improvement too big for the time. One change, measured well, beats three half-done ones.
- No references slide, or references copied in mixed styles. Use one style, usually IEEE; see the IEEE paper format for the reference style.
Frequently asked questions
How many slides should a base paper presentation have?
About 10 to 12 for a ten-minute slot: title, the problem, the paper’s details, its method, dataset, results, limitations, your proposed change, the plan and references. Your department’s time limit decides. The project synopsis format follows the same order in writing.
What should I say about the base paper’s results?
Show the paper’s own results table, name the dataset and the metric, and say what the numbers mean in one sentence. Do not claim results you have not reproduced. If you have run the code, show your reproduced number beside the paper’s and explain any gap.
How do I present the limitation of a base paper?
Pick one limitation you can test, such as a single dataset, no test on new conditions or a slow model, and show where the paper itself admits it or where the results reveal it. Then link it to the change you will make. Our guide to choosing a base paper explains how to spot a workable limitation.
Is a base paper presentation the same for B.Tech and M.Tech?
The slides are the same; the ending differs. A B.Tech review ends with the system you will build from the paper and how you will show it working. An M.Tech review ends with one measurable improvement and the comparison against the paper’s numbers, which becomes the core of the M.Tech thesis.
Can The Ultimate Project World help with a base paper presentation?
Yes. Send your base paper and review date in the free consultation. We help you understand the paper, reproduce its results where possible, choose the improvement and prepare the slides and answers, and can build the full project after the review. See our IEEE base paper projects.
